Structure Guided Design of Protein Biosensors for Phenolic Pollutants

Feb 22, 2017 - Gooding (Editor-in-Chief) , Kelley (Associate Editor) , Bakker (Associate Editor) , Long (Associate Editor) , Tao (Associate Editor) , ...
0 downloads 9 Views 2MB Size
Subscriber access provided by University of Newcastle, Australia

Article

Structure guided design of protein biosensors for phenolic pollutants Shamayeeta Ray, Santosh Panjikar, and Ruchi Anand ACS Sens., Just Accepted Manuscript • DOI: 10.1021/acssensors.6b00843 • Publication Date (Web): 22 Feb 2017 Downloaded from http://pubs.acs.org on February 26, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Sensors is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

Structure guided design of protein biosensors for phenolic pollutants Shamayeeta Ray 1, Santosh Panjikar 2,3 and Ruchi Anand*4,5 1

IITB-Monash Research Academy, Mumbai 400076, Maharashtra, India

2

Department of Biochemistry and Molecular Biology, Monash University, Victoria 3800, Australia

3

Australian Synchrotron, Victoria 3168, Australia

4

Department of Chemistry, Indian Institute of Technology Bombay, Mumbai 400076, Maharashtra, India

5

Wadhwani Research Center for Bioengineering, IIT Bombay, Mumbai 400076, India

*To whom correspondence may be addressed at: Dr. Ruchi Anand, Structural Biochemistry Laboratory, Department of Chemistry, Indian Institute of Technology Bombay, Mumbai 400076, Maharashtra, India. E-mail: [email protected]

ACS Paragon Plus Environment

ACS Sensors

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ABSTRACT Phenolic aromatic compounds are a major source of environmental pollution. Currently there are no in situ methods for specifically and selectively detecting these pollutants. Here, we exploit the nature’s biosensory machinery by employing Acinetobacter calcoaceticus NCIB8250 protein MopR, as a model system to develop biosensors for selective detection of a spectrum of these pollutants. The X-ray structure of the sensor domain of MopR was used as a scaffold for logicbased tunable biosensor design. By employing a combination of in silico structure guided approaches, mutagenesis and isothermal calorimetric studies, we were able to generate biosensor templates, that can selectively and specifically sense harmful compounds like chlorophenols, cresols, catechol and xylenols. Furthermore, the ability of native protein to selectively sense phenol as the primary ligand was also enhanced. Overall, this methodology can be extended as a suitable framework for development of a series of exclusive biosensors for accurate and selective detection of aromatic pollutants from real time environmental samples.

KEYWORDS: NtrC, MopR, phenol, catechol, xylenols, 3-chlorophenol, selective biosensor.

ACS Paragon Plus Environment

Page 2 of 21

Page 3 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

The phenolic class of xenobiotics like chlorophenols, catechol, dimethylphenols and cresols are major pollutants generated by oil, paper, tannery and several other industries1 and are listed as toxic and priority pollutants in the EPA and world pollution databases. Some of these compounds like bulkier dimethylphenols are resistant to biodegradation and have half-lives extending for several years in the environment2. These xenobiotics enter the soil and water bodies as waste and become dangerous if leaked into drinking water sources as they are extremely embryotoxic and carcinogenic in nature3. The current methods to detect these phenolic compounds rely on liquid-liquid extraction followed by gas chromatography, flame ionization detection and tandem mass spectrometry4,5. However, these methods are time consuming and require an elaborate instrumentation setup. Hence, there is a dire need to develop methodologies that can accurately, quickly and with high sensitivity detect these pollutants in real time. Devising chemical strategies that can differentiate between ligands like cresols, phenol, catechol and di-substituted phenols is a daunting task, mostly because these classes of compounds are by and large inert to detection. Hence, exploiting nature’s biosensory machinery that has been subject to evolutionary rigor, pose as a viable alternative. Prokaryotic soil bacteria like Pseudomonas sp. offer an excellent strategy towards developing cost effective biosensors for the said pollutants

6,7

. These bacteria possess special regulatory systems, which when faced with

adversity are capable of detecting specific pollutant molecules and degrading them8,9,10. The degradative pathways are either integrated into their host genome or are in the form of an entire gene cassette, helper plasmids11. The expression of these catabolic aromatic gene clusters is under the tight control of highly regulated transcription factors, like the nitrogen regulating class (NtrC) of proteins11-12. NtrC regulators like XylR, DmpR and MopR consists of three major domains, the Nterminal pollutant sensing domain (A), the AAA+ ATPase readout domain (C) and a Cterminally located DNA binding domain (D)13,14 (Figure 1A). The A and the C domains are connected by a flexible helical linker designated as B linker

15

(Figure 1A). Biochemical

experiments have shown that in the absence of the aromatic pollutants, the ATPase activity of the central AAA+ (C) domain remains repressed and activation occurs only upon binding of the target pollutants to the signal reception (A) domain16,17. Thus, it appears that this sensing design of these regulators can serve as an excellent platform for biosensor development, as both signal reception and readout domain exist within the same protein system.

ACS Paragon Plus Environment

ACS Sensors

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Over the past few decades, several attempts have been made to develop effective biosensors for an array of xenobiotics using this sub-class of regulators18,19. However, most of these targets were difficult to achieve due to lack of knowledge of the actual sensor determinants20,21. In the absence of any X-ray structure for this family, most of the mutations in the past were performed using various in silico models and via domain shuffling techniques20,21. Earlier this year, there was a breakthrough in the field where Ray et. al successfully solved the crystal structure of the signal sensing (A) and linker (B) domain of a protein from this subfamily, MopR (MopRAB), in complex with phenol and its derivatives22 (Figure 1B, PDB code: 5KBE). In parallel, Patil et.al. also solved the structure of the phenol sensing domain of PoxR, a close analog of MopR23. The crystal structure of MopRAB holds several surprises; although the protein is of bacterial origin, it resembles eukaryotic proteins and possess a fold similar to nitric oxide signaling and golgi transport enzymes that harbor long chain fatty acids molecules24. The MopR protein also contains a novel zinc binding motif that does not play any role in sensing, however, imparts overall structural stability.

The structure reveals that the pollutant is nested in a

hydrophobic pocket lined by residues like phenylalanine, isoleucine, alanine and tyrosine with the phenolic ligand being anchored by the indole nitrogen atom of W134 and the imidazole nitrogen atom of H106 ( Figure 1C ). Moreover, supporting isothermal calorimetry (ITC) data shows that native MopR binds only small phenolic ligands and is unable to tolerate much variation in ligand architecture ( Figure 5B, Table S2 )22. Armed with the structural information, here we used the ligand binding pocket as a template to expand its sensing repertoire and to create several specific biosensor models for selective sensing of toxic environmental pollutants like hazardous air pollutant catechol, highly corrosive pollutants 3-chlorphenol (3-cp) and cresols, toxic aquatic pollutants xylenols and priority pollutant phenol. The substrate scope was altered by employing targeted mutagenesis as the primary tool. The design of the mutations was first refined by in silico mutagenesis along with docking studies and a recombinant MopR-aromatic pollutant pair for each mutation was identified. The in silico results were corroborated by carrying out the select mutations, purifying the proteins and performing ITC based binding profile characterizations of the different pollutant-MopRAB mutant combinations. The biosensor design was further substantiated by augmenting the ATPase readout (C) domain to different MopRAB variants and creating MopRABC mutants that served as a sensing chimera for direct pollutant readout down to low ppm levels.

ACS Paragon Plus Environment

Page 4 of 21

Page 5 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

Together, this structure based approach provides enhanced selectivity and represents an appropriate scaffold for development of a series of selective and exclusive biosensors for a host of hazardous environmental pollutants.

EXPERIMENTAL SECTION Docking Studies. Based on the detailed analysis of the different aromatic pollutant bound structures of the signal sensing (A) domain along with part of the linker region (B) of MopR (MopRAB)22, various site-specific mutations of the two major phenol binding residues H106 and W134 were designed in silico (Figure 1A). Docking experiments were performed using AutoDock version 4.225 with the mutated constructs to test for any alterations in specificity and selectivity of the substrate scope of MopRAB (Table S1). The mutations included - (i) Alanine substitutions, MopRABHA (H106A substitution), MopRABWA (W134A substitution) and MopRABWHA (W134A-H106A double mutation), (ii) Asparagine substitution MopRABHN (H106N substitution) and (iii) Tyrosine substitution of H106, MopRABHY (H106Y substitution) and they were docked with the following select phenolic pollutants - phenol, o-cresol, m-cresol, catechol, 3-chlorophenol(3-cp) and 3,4-dimethylphenol (3,4-dmp) (Figure 5A, Table S1) (Detailed in Supporting Materials and Methods). DNA manipulations, over expression and purification of the recombinant proteins. To validate the docking results, the same set of in silico mutations were experimentally incorporated into the native MopRAB construct from Acinetobacter calcoaceticus NCIB8250 that was previously cloned into modified pET vector22. Further, to test the translational biosensing ability of the different MopRAB sensor designs, a longer construct of the native MopR (MopRABC) consisting of both the signal sensing (A and B) and the readout ATPase (C) domain, was cloned into modified pET vector (Figure 1A). The native MopRABC was then used as a template to perform following mutations within the bigger construct - MopRABCHY (H106Y substitution), MopRABCHN (H106N substitution) and MopRABCHA (H106A substitution). All the point mutations in MopRAB and MopRABC were performed by employing standard site-directed mutagenesis protocol using the "site-directed mutagenesis kit" from Kapa biosystems. The native as well as the mutated protein constructs were expressed as His-tag fusion proteins and were purified using standard His-tagged affinity purification protocol (Detailed in Supporting Materials and Methods).

ACS Paragon Plus Environment

ACS Sensors

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Ligand binding experiment using ITC. In order to validate the in silico affinity of different MopRAB mutants towards selective phenolic pollutants, in-vitro ligand binding experiments were performed using MicroCal iTC200 (GE Healthcare, WI, USA) (Detailed in Supporting Materials and Methods, Table S2). Colorimetric ATPase assay design. The biosensing ability of native and mutated MopRABC towards a variety of aromatic pollutants were tested using malachite green based colorimetric ATPase assay26 and monitored spectrophotometrically at 630 nm ( Details of assay protocol provided in Supporting Materials and Methods).

RESULTS AND DISCUSSION In order to create modified sensor designs by employing the structure as a template, a series of pollutant targets ranging from phenol, xylenols and catechol were chosen. The first set of biosensors were designed to be specific towards ortho and meta substituted phenolic ligands like o-cresol and 3-cp respectively. The native structure clearly reveals that both these pollutants snugly fit into the binding pocket in a fixed conformation with little room for re-orientation or for accepting any extra side groups22. Therefore, to alter the substrate scope in favor of meta and ortho directed aromatics, one strategy that was undertaken was to create space in the pocket. To do so only one of the phenolic anchors H106 or W134 was retained and other was mutated to an alanine residue. This would ensure additional free movement that can enable the differently directed ligands to be selectively accommodated. To test this hypothesis, in silico mutations involving single alanine substitution of the key sensor residues - MopRABHA (H106A) and MopRABWA (W134A) were constructed. Docking MopRABHA and MopRABWA with a subset of meta and ortho substituents revealed that MopRABHA preferred meta-substituted effectors, whereas MopRABWA favored ortho directed compounds (Figure 2). This is because in MopRABHA, the meta directed compounds are able to flip in the active site and the phenolic OH now instead of anchoring with W134, forms strong hydrogen bond with the carbonyl group of residue A162. The binding affinity in this conformation is likely to be retained as the ligand still fits into the pocket in a favorable orientation with its meta directed group accommodating itself in the space created by the lack of histidine residue (Figure 2A). Similarly, ortho-substituted ligands prefer MopRABWA mutant as again, due to the shape complementarity, the ligand is able to flip and interact with the backbone of A162 through hydrogen bonding interactions with the main chain

ACS Paragon Plus Environment

Page 6 of 21

Page 7 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

carbonyl group without any steric clash (Figure 2E). Docking also shows that the MopRABWA mutation is unfavorable towards meta directed ligands as these compounds are neither able to flip in the pocket, due to a steric clash of the meta-directed substituent of the flipped ligand with H106, nor maintain the original conformation (Figure 2C). Instead, the meta-oriented compound occupies the empty space created by the W134A mutation adopting a conformation approximately 7 Å away from the original optimal position. Similarly, MopRABHA docked with o-cresol shows that it does not favor ligands with ortho substitutions as the phenolic OH group comes in close proximity of another electron donating S166-OH (Figure 2G), an energetically unfavorable scenario. Experimental results with purified mutants (Supporting Materials and Methods) corroborate the in silico predictions. The binding affinity (Kd) of these mutants showed that MopRABHA prefers meta-substituted phenols over their ortho directed counterparts by over 10 fold (Figure 2B, 2H, 5B, S1, Table S2). An opposite trend in binding affinity was observed for the MopRABWA mutation where ortho oriented compounds bind with much higher affinity than their meta directed counterparts (Figure 2D, 2F, 5C, S2, Table S2). Hence, engineering the binding pocket with these alanine substitutions have led to creation of sensor frameworks with enhanced selectivity towards ortho- or meta- substituted phenolic pollutants . Our next goal was to generate an exclusive sensor for catechol, a water soluble, volatile skin and eye irritant, which act as a central nervous system depressant and can cause hypertension and convulsions. Since catechol can easily enter the water sources, there is a dire need to monitor its levels. Our ortho directed study revealed that catechol being bulky with two hydroxyl groups might fit better in a flipped orientation. To facilitate the flip, the suggested design would have the histidine position blocked with an alternative amino acid with the anchor properties obliterated. The tyrosine residue having an aromatic hydroxyl group, fits the above mentioned parameters, hence an in silico MopRABHY (H106Y) mutation was performed. Docking results were extremely encouraging; the flipped catechol adopted a favorable conformation without any steric clash (Figure 3C). In this conformation, one of its OH moieties makes a strong hydrogen bond with the carbonyl group of A162 and the other OH interacts with Y176 leading to further stabilization (Figure 3A). Docking with all other phenol derivatives however, resulted in unfavorable orientations producing no logical contacts with the MopRABHY pocket residues (Figure 3D). Experimental validation of MopRABHY again confirms the predictions and reveals that phenol and most of the other phenolic effectors, exhibit extremely poor affinity towards

ACS Paragon Plus Environment

ACS Sensors

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 21

MopRABHY (Figure 5G, S3, Table S2). The only exception was catechol that showed substantial affinity (Kd of 5.07 ± 0.12 µM) (Figure 3B, 5G, Table S2). These results indicate that MopR sensor binding pocket is extremely sensitive to single point changes, and a sole mutation can completely alter the binding profile of the sensor protein. Thus the MopRABHY mutant is selective for stabilization of catechol, providing a framework to design an exclusive catechol sensor. Xylenols (bulkier phenols) like dimethylphenols, are long lasting aquatic pollutants that enter the environment through processes like production of phenolic resins commonly used for poly(p-phenylene oxide), antioxidants and varnishes production.

In order to develop

selective sensors for these xylenols, the current aim was to increase the pocket size while retaining the phenolic anchor. It was noticed from studies on the previous sensors (MopRABHY, MopRABHA) that an effective strategy was to use A162 as a phenolic anchor and flip the alcohol. Hence, an adaptive design with increased pocket volume was created in silico that involves a double alanine substitution of both W134 and H106 residues generating the MopRABWHA (W134A-H106A) construct design. The docking studies with one of the xylenols, 3,4dimethylphenol(3,4-dmp), shows that it flips in the MopRABWHA pocket in an orientation as predicted, with the OH group making strong hydrogen bonding interaction with the main chain carbonyl group of A162. The double alanine substitution creates sufficient space in the pocket to accommodate both the methyl groups of 3,4-dmp in favorable conformation without any steric clash leading to an overall stabilized state (Figure 4A). In contrast, single mutants like MopRABHA , MopRABWA and MopRABHN (H106N) could not create enough space to allow a proper anchoring of 3,4-dmp (Figure 4C, S4A, S4C). The ITC results clearly reasserts the observations from docking as 3,4-dmp has the highest affinity for the MopRABWHA mutant (Kd of 4.99 ± 0.06 µM) (Figure 4B, 5E, Table S2) and shows extremely poor affinity or almost no binding towards the other single mutants (Figure 4D, S4B, S4D, Table S2). MopRABWHA also shows some affinity towards other meta-directed bulkier phenol derivatives like 3-cp and mcresol (Figure S5, 5E, Table S2) but their affinity is much less compared to 3,4-dmp. Based on all these observations, it can be inferred that MopRABWHA can serve as a model construct for selective sensing of bulkier toxic aromatic pollutants. These results reflect that logic-based tweaking of the MopR binding pocket can indeed help in sensing new effectors which are usually weak binders of the native protein.

ACS Paragon Plus Environment

Page 9 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

Although the native protein senses phenol with high affinity, it exhibited some level of promiscuity. Hence, we shifted our focus on improving the phenol sensing design of the native MopR. This will aid in exclusive detection of phenol without any background signal from other alcohols. To facilitate this, a design tailored to conformationally constrict the phenolic anchor will be the most suited. Therefore, the anchor H106 was replaced with an asparagine residue which has a similar size and pKa value (approximately ~ 9) like the delta nitrogen of histidine residue, but adopts specific rotamers in the pocket. In silico H106N substitution (MopRABHN) shows that only phenol can be effectively docked in the MopRABHN binding pocket as compared to all other phenol derivatives (Figure 4E, 4G, S6A, S6C). Docking further shows that the OH group of the phenol retained the hydrogen bond with W134 along with formation of a new bond with N106 leading to an overall stabilization of the ligand (Figure 4E). However, for o-cresol (having the methyl group at the ortho positon), it was observed that the ligand rotated in a manner where it could make hydrogen bond with only W134 and not N106 leading to weaker interactions (Figure S6A). In case of m-cresol and catechol, both the ligands flipped within the MopRABHN pocket and their OH groups could make only weak hydrogen bonds with Y176-OH and S166-OH leading to overall destabilization (Figure 4G, S6C). In vitro ligand binding experiments of MopRABHN validated the docking results, where only phenol exhibited high affinity (Kd of 1.59 ± 0.23 µM) (Figure 4F) and all the other phenol derivatives showed poor affinity towards MopRABHN (Figure 4H, S6B, S6D,5F, Table S2). Though the affinity of MopRABHN towards phenol is slightly reduced as compared to the native protein (Kd of 0.46 ± 0.06) (Table S2) but the advantage of this mutation lies in the fact that it can now selectively sense phenol over other pollutants, a property which is lacking in native MopRAB (Figure 5B, 5F). These observations highlight the importance of correct choice of amino acids in the active site and assert the fact that substitution of any hydrophilic amino acid with another capable of hydrogen bonding is not sufficient to anchor a wide spectrum of phenolic pollutants within the MopR pocket. To demonstrate that the designed model systems can be directly translated as biosensor units, longer constructs of MopR consisting of both the signal sensing(A) (pollutant binding) and the readout(C) (ATP hydrolysis) domain (Figure 1A) for the native (MopRABC) as well as three single mutants, MopRABCHA (H106A, corresponding to the meta-phenol sensor model), MopRABCHY ( (H106Y, corresponding to the catechol sensor model) and MopRABCHN (

ACS Paragon Plus Environment

ACS Sensors

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 21

(H106N, corresponding to the phenol sensor model) were constructed and purified. The ATPase activity of both native and mutated MopRABC which gets activated on target pollutant binding to the sensing domain, was quantified by colorimetric estimation26. Assay with each protein construct was performed with a host of simulated wastewater samples, each containing a different aromatic pollutant as the primary contaminant (Figure 5A). To confirm viability of the design, each of the pollutant's concentration used for testing were set at 10µM (corresponding to 0.94 ppm for phenol, 1.1 ppm for catechol, 1.08 ppm for m-cresol and o-cresol, 1.28 ppm for 3cp and 1.22 ppm for 3,4-dmp ) which lies below the approximate environmental risk limits as per Occupational Safety and Health Administration (OSHA). Results show that each MopRABC mutant protein exhibited significant ATPase activity only towards those compounds for which that particular MopRAB construct has high affinity (Figure 6B-D), thereby validating our docking and ITC studies. Hence, native MopRABC displayed a broader sensing spectrum ( Figure 6A), whereas the mutants behaved as selective biosensors (Figure 6B-D). Similar translational biosensing responses are expected for the other sensor designs. Therefore, these selective sensors pave the way towards design of biosensing tools to gauge levels of particular contaminants in wastewater environmental samples and to categorize the type of pollutants present in them.

CONCLUSION In summary, the in silico as well as experimental analysis of various phenolic pollutants with the different MopR variants highlight the fact that structure guided protein engineering of the binding pocket can help generate selective biosensors that can be designed to possess enhanced ability to target one or more hazardous aromatic pollutants. The efficient sensor designs obtained from this work include an exclusive phenol sensor (MopRABHN), ortho-phenol sensor (MopRABWA), meta-phenol sensor (MopRABHA), xylenol sensor (MopRABWHA) and a catechol sensor (MopRABHY) (Figure 5C-G). An advantage of this work is that the ligand binding pocket of the same template protein can be engineered according to the target requirements, which makes it an economical and efficient approach. At present, our tested selective biosensors can sense the contaminants below the estimated risk limits as per OSHA. In order to achieve even higher sensitivity of these selective sensors for pollutant detection, further efforts are underway towards construction and quantitative detection of these aromatic pollutants from real time

ACS Paragon Plus Environment

Page 11 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

environmental samples. This study is a stepping stone towards efficient bioremediation of target aromatic pollutants.

Supporting Information The Supporting Information is available free of charge on the ACS Publications Website. Supporting Information includes Supporting Materials and Methods, Supporting References, six figures and two tables.

Author Information Corresponding Author *E-mail: [email protected]. Author Contributions S.R., S.P., R.A. designed research; S.R. performed research; R.A. contributed new reagents/analytic tools; S.R., S.P., R.A. analyzed data; and S.R., S.P., R.A. wrote the paper. Notes The authors declare no competing financial interest.

Acknowledgements This work was funded by DST, Government of India (Grant Numbers EMR/2015/002121 and DST/TM/WTI/2K16/252), Wadhwani Research Center for Bioengineering (WRCB), IIT Bombay, India and Australia-India Council (AIC) grant.

References 1.

Keith, L. H., The Source of U.S. EPA's Sixteen PAH Priority Pollutants. Polycycl. Aromat. Comp.

2015, 35, 147-160. 2.

Golovleva, L. A.; Aharonson, N.; Greenhalgh, R.; Sethunathian, N.; Vonk, J. W., The role and

limitations of microorganisms, in the conversion of xenobiotics (IUPAC reports on pesticides no. 27). Pure Appl. Chem. 1990, 62 (2), 351-364. 3.

Abdel-Shafy, H. I.; Mansour, M. S. M., A review on polycyclic aromatic hydrocarbons: Source,

environmental impact, effect on human health and remediation. Egypt. J. Petrol. 2016, 25 (1), 107-123.

ACS Paragon Plus Environment

ACS Sensors

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

4.

Page 12 of 21

Mahugo Santana, C.; Sosa Ferrera, Z.; Esther Torres Padrón, M.; Juan Santana Rodríguez, J.,

Methodologies for the extraction of phenolic compounds from environmental samples: new approaches. Molecules 2009, 14 (1), 298-320. 5.

Wille, K.; De Brabander, H. F.; Vanhaecke, L.; De Wulf, E.; Van Caeter, P.; Janssen, C. R., Coupled

chromatographic and mass-spectrometric techniques for the analysis of emerging pollutants in the aquatic environment. Trends Anal. Chem. 2012, 35, 87-108. 6.

Shingler, V., Integrated regulation in response to aromatic compounds: from signal sensing to

attractive behaviour. Environ. Microbiol. 2003, 5 (12), 1226-1241. 7. Diaz, E.; Prieto, M. A., Bacterial promoters triggering biodegradation of aromatic pollutants. Curr. Opin. Biotech. 2000, 11 (5), 467-75. 8. Timmis, K. N.; Pieper, D. H., Bacteria designed for bioremediation. Trends Biotech. 1999, 17 (5), 201-204. 9. Tropel, D.; van der Meer, J. R., Bacterial transcriptional regulators for degradation pathways of aromatic compounds. Microbiol. Mol. Biol. Rev. 2004, 68 (3), 474-500. 10. Gerischer, U., Specific and global regulation of genes associated with the degradation of aromatic compounds in bacteria. J. Mol. Microbiol. Biotechnol. 2002, 4 (2), 111-121. 11. Worsey, M. J.; Franklin, F. C.; Williams, P. A., Regulation of the degradative pathway enzymes coded for by the TOL plasmid (pWWO) from Pseudomonas putida mt-2. J. Bacteriol. 1978, 134 (3), 757764. 12. Shingler, V., Signal sensing by σ54-dependent regulators: derepression as a control mechanism. Mol. Microbiol. 1996, 19 (3), 409-416. 13. North, A. K.; Klose, K. E.; Stedman, K. M.; Kustu, S., Prokaryotic enhancer-binding proteins reflect eukaryote-like modularity: the puzzle of nitrogen regulatory protein C. J. Bacteriol. 1993, 175 (14), 42674273. 14. Bush, M.; Dixon, R., The Role of Bacterial Enhancer Binding Proteins as Specialized Activators of σ(54)-Dependent Transcription. Microbiol. Mol. Biol. Rev. 2012, 76 (3), 497-529. 15. Wootton, J. C.; Drummond, M. H., The Q-linker: a class of interdomain sequences found in bacterial multidomain regulatory proteins. Protein Eng., Des. Sel. 1989, 2, 535−543. 16. O'Neill, E.; Ng, L. C.; Sze, C. C.; Shingler, V., Aromatic ligand binding and intramolecular signalling of the phenol-responsive σ54-dependent regulator DmpR. Mol. Microbiol. 1998, 28 (1), 131141. 17. Ng, L. C.; O'Neill, E.; Shingler, V., Genetic Evidence for Interdomain Regulation of the Phenolresponsive 54-dependent Activator DmpR. J. Biol. Chem. 1996, 271 (29), 17281-17286.

ACS Paragon Plus Environment

Page 13 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

18. Kim, M. N.; Park, H. H.; Lim, W. K.; Shin, H. J., Construction and comparison of Escherichia coli whole-cell biosensors capable of detecting aromatic compounds. J. Microbiol. Meth. 2005, 60 (2), 235245. 19. Gupta, S.; Saxena, M.; Saini, N.; Mahmooduzzafar; Kumar, R.; Kumar, A., An effective strategy for a whole-cell biosensor based on putative effector interaction site of the regulatory DmpR protein. PloS one 2012, 7 (8), e43527. 20. Skärfstad, E.; O'Neill, E.; Garmendia, J.; Shingler, V., Identification of an Effector Specificity Subregion within the Aromatic-Responsive Regulators DmpR and XylR by DNA Shuffling. J. Bacteriol. 2000, 182 (11), 3008-3016. 21. Devos, D.; Garmendia, J.; Lorenzo, V. d.; Valencia, A., Deciphering the action of aromatic effectors on the prokaryotic enhancer-binding protein XylR: a structural model of its N-terminal domain. Environ. Microbiol. 2002, 4 (1), 29-41. 22. Ray, S.; Gunzburg, M. J.; Wilce, M.; Panjikar, S.; Anand, R., Structural Basis of Selective Aromatic Pollutant Sensing by the Effector Binding Domain of MopR, an NtrC Family Transcriptional Regulator. ACS Chem. Biol. 2016, 11 (8), 2357–2365. 23. Patil, V. V.; Park, K. H.; Lee, S. G.; Woo, E., Structural Analysis of the Phenol-Responsive Sensory Domain of the Transcription Activator PoxR. Structure 2016, 24 (4), 624-30. 24. Turnbull, A. P.; Kümmel, D.; Prinz, B.; Holz, C.; Schultchen, J.; Lang, C.; Niesen, F. H.; Hofmann, K.-P.; Delbrück, H.; Behlke, J.; Müller, E.-C.; Jarosch, E.; Sommer, T.; Heinemann, U., Structure of palmitoylated BET3: insights into TRAPP complex assembly and membrane localization. EMBO J. 2005, 24 (5), 875-884. 25. Morris, G. M.; Huey, R.; Lindstrom, W.; Sanner, M. F.; Belew, R. K.; Goodsell, D. S.; Olson, A. J., AutoDock4 and AutoDockTools4: Automated docking with selective receptor flexibility. J. Comput. Chem. 2009, 30(16), 2785-91. 26. Baykov, A. A.; Evtushenko, O. A.; Avaeva, S. M., A malachite green procedure for orthophosphate determination and its use in alkaline phosphatase-based enzyme immunoassay. Anal. Biochem. 1988, 171 (2), 266-70.

FIGURE LEGENDS Figure 1. Structural insights into the pollutant sensing domain of MopR. (A) Domain organization of MopR, an NtrC family regulator22. Site-specific mutations have been carried out on the signal sensing anchors (listed in red) in the pollutant sensing (A) domain (coloured light blue) of MopR . (B) Crystal structure of the pollutant sensing (A) and linker (B) domain of MopR (MopRAB) nesting a zinc atom (in orange) and a bound phenol (in green)22 [ Panels (A) and (B) have been adapted in part from 'Ray, S.; Gunzburg, M. J.; Wilce, M.; Panjikar, S.; Anand, R., Structural Basis of Selective Aromatic

ACS Paragon Plus Environment

ACS Sensors

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 21

Pollutant Sensing by the Effector Binding Domain of MopR, an NtrC Family Transcriptional Regulator. ACS Chem. Biol. 2016, 11 (8), 2357– 2365'. Copyright ©2016 American Chemical Society]

(C) Magnified stereo view of the pollutant binding pocket of MopRAB as

obtained from the phenol-bound crystal structure. The two sensor residues W134 and H106 anchoring the phenolic OH are in firebrick and the other ligand-binding residues are in cyan. Oxygen and nitrogen atoms are in red and blue respectively. Figure 2. Selective sensing of ortho and meta directed phenolic pollutants. The panels represent docked ligands and the ITC curves for the following MopRAB mutants (A-B) 3-cp with MopRABHA, (C-D) 3-cp with MopRABWA, (E-F) o-cresol with MopRABWA and (G-H) o-cresol with MopRABHA. Pocket residues are in cyan and mutated residues in firebrick. Oxygen and nitrogen atoms are in red and blue respectively. The ITC data corresponding to each docking experiment is given in the panels below. ITC data were fit using one set of sites model and the thermodynamic parameters are given in Table S2. All the Kd values represented in the figure are in µM. Figure 3. Selective catechol sensor design. Panel (A) and (B) represent docked catechol and the ITC curve for MopRABHY mutant respectively. Panel (C) and (D) shows the surface representation of the MopRABHY binding pocket with catechol and 3,4-dmp respectively. Carbon atoms of all the phenolic ligands are colored orange, pocket residues are in cyan and mutated residues in firebrick. The surface representation of the tyrosine side chain is in deep blue. Oxygen and nitrogen atoms are in red and blue respectively. ITC data was fit using one set of sites model and the thermodynamic parameters are given in Table S2. The Kd value represented in the figure are in µM. Figure 4. Selective sensing of xylenols and phenol. The panels represent docked ligands and the ITC curves for the following MopRAB mutants- (A-B) 3,4-dmp with MopRABWHA, (C-D) 3,4-dmp with MopRABWA, (E-F) phenol with MopRABHN and (G-H) m-cresol with MopRABHN. Pocket residues are in cyan and mutated residues in firebrick. Oxygen and nitrogen atoms are in red and blue respectively. The ITC data corresponding to each docking experiment is given in the panels below. ITC data were fit using one set of sites model and the thermodynamic parameters are given in Table S2. All the Kd values represented in the figure are in µM. Figure 5. Binding affinity of MopRAB and its mutants towards different phenol derivatives. (A) represents structures of the aromatic compounds. Panel (B) - (G) represent the affinity of various MopRAB constructs towards different phenol derivatives based on the Kd values obtained from the ITC data of each ligand-protein run (Table S2). Y-axis in each bar diagram represents percent (%) binding affinity (computed based on 1/ Kd) for each ligand. The compounds for whom Kd values are "Not determinable" have been assigned value of 1 and represent those ITC runs for which some heat change was observed but the data could not be fit using any standard binding curve models and hence, their thermodynamic parameters couldn't be computed. "No binding" refers to those ITC runs where there was negligible heat change and hence, have been assigned value of 0 in the bar plots shown above. Figure 6. Translational biosensing ability of various selective sensor designs of MopR. Panels represent percent (%) ATPase activity of native (A) and mutated (B-D) MopRABC constructs (comprising of the pollutant sensing (A) and readout ATPase(C) domain) towards select pollutants, tested using a colorimetric ATPase assay26.

ACS Paragon Plus Environment

Page 15 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

Figure 1. Structural insights into the pollutant sensing domain of MopR. (A) Domain organization of MopR, an NtrC family regulator22. Site-specific mutations have been carried out on the signal sensing anchors (listed in red) in the pollutant sensing (A) domain (coloured light blue) of MopR . (B) Crystal structure of the pollutant sensing (A) and linker (B) domain of MopR (MopRAB) nesting a zinc atom (in orange) and a bound phenol (in green)22 [ Panels (A) and (B) have been adapted in part from 'Ray, S.; Gunzburg, M. J.; Wilce, M.; Panjikar, S.; Anand, R., Structural Basis of Selective Aromatic Pollutant Sensing by the Effector Binding Domain of MopR, an NtrC Family Transcriptional Regulator. ACS Chem. Biol. 2016, 11 (8), 2357–2365'. Copyright ©2016 American Chemical Society] (C) Magnified stereo view of the pollutant binding pocket of MopRAB as obtained from the phenol-bound crystal structure. The two sensor residues W134 and H106 anchoring the phenolic OH are in firebrick and the other ligand-binding residues are in cyan. Oxygen and nitrogen atoms are in red and blue respectively. 77x47mm (300 x 300 DPI)

ACS Paragon Plus Environment

ACS Sensors

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 2. Selective sensing of ortho and meta directed phenolic pollutants. The panels represent docked ligands and the ITC curves for the following MopRAB mutants (A-B) 3-cp with MopRABHA, (C-D) 3-cp with MopRABWA, (E-F) o-cresol with MopRABWA and (G-H) o-cresol with MopRABHA. Pocket residues are in cyan and mutated residues in firebrick. Oxygen and nitrogen atoms are in red and blue respectively. The ITC data corresponding to each docking experiment is given in the panels below. ITC data were fit using one set of sites model and the thermodynamic parameters are given in Table S2. All the Kd values represented in the figure are in µM. 98x54mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 16 of 21

Page 17 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

Figure 3. Selective catechol sensor design. Panel (A) and (B) represent docked catechol and the ITC curve for MopRABHY mutant respectively. Panel (C) and (D) shows the surface representation of the MopRABHY binding pocket with catechol and 3,4-dmp respectively. Carbon atoms of all the phenolic ligands are colored orange, pocket residues are in cyan and mutated residues in firebrick. The surface representation of the tyrosine side chain is in deep blue. Oxygen and nitrogen atoms are in red and blue respectively. ITC data was fit using one set of sites model and the thermodynamic parameters are given in Table S2. The Kd value represented in the figure are in µM. 119x110mm (300 x 300 DPI)

ACS Paragon Plus Environment

ACS Sensors

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 4. Selective sensing of xylenols and phenol. The panels represent docked ligands and the ITC curves for the following MopRAB mutants- (A-B) 3,4-dmp with MopRABWHA, (C-D) 3,4-dmp with MopRABWA, (EF) phenol with MopRABHN and (G-H) m-cresol with MopRABHN. Pocket residues are in cyan and mutated residues in firebrick. Oxygen and nitrogen atoms are in red and blue respectively. The ITC data corresponding to each docking experiment is given in the panels below. ITC data were fit using one set of sites model and the thermodynamic parameters are given in Table S2. All the Kd values represented in the figure are in µM. 97x53mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 18 of 21

Page 19 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

Figure 5. Binding affinity of MopRAB and its mutants towards different phenol derivatives. (A) represents structures of the aromatic compounds. Panel (B) - (G) represent the affinity of various MopRAB constructs towards different phenol derivatives based on the Kd values obtained from the ITC data of each ligandprotein run (Table S2). Y-axis in each bar diagram represents percent (%) binding affinity (computed based on 1/ Kd) for each ligand. The compounds for whom Kd values are "Not determinable" have been assigned value of 1 and represent those ITC runs for which some heat change was observed but the data could not be fit using any standard binding curve models and hence, their thermodynamic parameters couldn't be computed. "No binding" refers to those ITC runs where there was negligible heat change and hence, have been assigned value of 0 in the bar plots shown above. 121x98mm (300 x 300 DPI)

ACS Paragon Plus Environment

ACS Sensors

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 6. Translational biosensing ability of various selective sensor designs of MopR. Panels represent percent (%) ATPase activity of native (A) and mutated (B-D) MopRABC constructs (comprising of the pollutant sensing (A) and readout ATPase(C) domain) towards select pollutants, tested using a colorimetric ATPase assay26. 127x148mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 20 of 21

Page 21 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Sensors

TOC 47x32mm (300 x 300 DPI)

ACS Paragon Plus Environment